With the Internet technology improving constantly, cloud storage technologies show prominent advantages in data storage, users can store data in cloud storage systems. For example, in the field of video surveillance, a large number of video data are often stored in data cloud storage systems implemented by cloud storage technologies. Currently, most data cloud storage systems are distributed systems adopting a client-server model. A master node (i.e. a scheduling management service center) can manage various slave nodes (i.e. storage servers) in the system. When the master node fails, the entire data cloud storage system will not function properly. Obviously, such data cloud storage systems are very sensitive to the running state of the master node.
In order to solve the problem that the entire data cloud storage system cannot run normally due to the failure of a master node, a hot standby method is provided in the prior art. There are two master nodes (a master node A and a master node A′) running at the same time in the system. If the master node A which is currently running fails, it immediately switches to the standby master node A′, thus ensuring that the entire data cloud storage system can continue under normal operation.
It can be seen that the above method can guarantee, to a certain extent, the normally operation of the data cloud storage system. However, this method merely replaces the master node A by the master node A′ when the master node A fails, to ensure the normal operation of that system. In the case that the master node A has not yet returned to a normal running state, and the standby master node fails too, the data cloud storage system will not function properly as well. Obviously, such data cloud storage systems are also very sensitive to the running state of the master node or the standby master node.